encoding csv

dusha.py

@@ -46,7 +46,7 @@ class Dusha(datasets.GeneratorBasedBuilder):
     def _generate_examples(self, audio_files: Union[str, os.PathLike], metadata: str):
         examples = dict()
 
-        with open(metadata, encoding="cp1251") as f:
+        with open(metadata, encoding="utf-8") as f:
             for row in f:
                 data = row.split(",")
                 audio_path = data[0]